Pol II–Expressed shRNA Knocks Down Sod2
Gene Expression and Causes Phenotypes
of the Gene Knockout in Mice
Xu-Gang Xia1[, Hongxia Zhou1[, Enrique Samper2, Simon Melov2, Zuoshang Xu1,3,4*
1 Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, Massachusetts, United States of America, 2 Buck
Institute for Age Research, Novato, California, United States of America, 3 Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts, United States
of America, 4 Neuroscience Program, University of Massachusetts Medical School, Worcester, Massachusetts, United States of America
RNA interference (RNAi) has been used increasingly for reverse genetics in invertebrates and mammalian cells, and has
the potential to become an alternative to gene knockout technology in mammals. Thus far, only RNA polymerase III
(Pol III)–expressed short hairpin RNA (shRNA) has been used to make shRNA-expressing transgenic mice. However,
widespread knockdown and induction of phenotypes of gene knockout in postnatal mice have not been demonstrated.
Previous studies have shown that Pol II synthesizes micro RNAs (miRNAs)—the endogenous shRNAs that carry out
gene silencing function. To achieve efficient gene knockdown in mammals and to generate phenotypes of gene
knockout, we designed a construct in which a Pol II (ubiquitin C) promoter drove the expression of an shRNA with a
structure that mimics human miRNA miR-30a. Two transgenic lines showed widespread and sustained shRNA
expression, and efficient knockdown of the target gene Sod2. These mice were viable but with phenotypes of SOD2
deficiency. Bigenic heterozygous mice generated by crossing these two lines showed nearly undetectable target gene
expression and phenotypes consistent with the target gene knockout, including slow growth, fatty liver, dilated
cardiomyopathy, and premature death. This approach opens the door of RNAi to a wide array of well-established Pol II
transgenic strategies and offers a technically simpler, cheaper, and quicker alternative to gene knockout by
homologous recombination for reverse genetics in mice and other mammalian species.
Citation: Xia XG, Zhou H, Samper E, Melov S, Xu Z (2006) Pol II–expressed shRNA knocks down Sod2 gene expression and causes phenotypes of the gene knockout in mice.
PLoS Genet 2(1): e10.
Gene knockout by homologous recombination has been
instrumental in investigating gene functions in mammals. It
has been used to reveal gene functions in normal as well as in
pathogenic pathways in vivo, and to generate models for
many genetic disorders. However, the technical complexity,
lengthy process, and high cost have limited its broad
application. This is particularly problematic considering the
fact that of the ;30,000 currently known mouse genes, only
10% have been knocked out, and even fewer are readily
accessible by the research community . Furthermore, the
knockout technology, in most instances, generates animals
with ;50% or 0% expression of the target gene, and it is
difficult to create graded hypomorphic models, which may be
necessary for modeling some diseases. In addition, the
knockout technology is not well established in other
mammalian species, consequently limiting the investigation
of gene functions and the development of disease models in
other mammalian species. These limitations may be over-
come by RNA interference (RNAi) technology [1,2].
RNAi is a widely conserved mechanism in eukaryotes .
Triggered by double-stranded RNA (dsRNA) in cells, RNAi
destroys the target RNA that shares sequence homology with
the dsRNA . The mechanism of RNAi is not fully under-
stood. A simplified model based mainly on data from
Drosophila has the following steps: First, Dicer, an enzyme of
the RNase III family, initiates ATP-dependent fragmentation
of long dsRNA into 21- to 25-nucleotide double-stranded
fragments, called small interfering RNAs (siRNAs). Second,
the siRNA duplexes bind the proteins Dicer and R2D2, which
facilitate the formation of a siRNA/multiprotein complex
called RNA-induced silencing complex (RISC) loading com-
plex. Third, the siRNA duplex in the RISC loading complex
unwinds to form an active RISC that contains a single-
stranded RNA (called the guide strand). Fourth, the RISC
recognizes the target RNA by Watson–Crick base pairing with
the guide strand and cleaves the target RNA. Finally, the RISC
releases its cleaved product and goes on to catalyze a new
cycle of target recognition and cleavage .
In differentiated mammalian cells, long dsRNA activates
RNA-dependent protein kinase PKR and type I interferon
Editor: David Beier, Harvard Medical School, United States of America
Received October 13, 2005; Accepted December 14, 2005; Published January 27,
A previous version of this article appeared as an Early Online Release on December
14, 2005 (DOI: 10.1371/journal.pgen.0020010.eor).
Copyright: ? 2006 Xia et al. This is an open-access article distributed under the
terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author
and source are credited.
Abbreviations: dsRNA, double-stranded RNA; EGFP, enhanced green fluorescent
protein; miRNA, micro RNA; Pol, RNA polymerase; RISC, RNA-induced silencing
complex; RNAi, RNA interference; SDH, succinate dehydrogenase; shRNA, short
hairpin RNA; siRNA, small interfering RNA
* To whom correspondence should be addressed. E-mail: zuoshang.xu@umassmed.
[ These authors contributed equally to this work.
PLoS Genetics | www.plosgenetics.org January 2006 | Volume 2 | Issue 1 | e100073
response, which leads to a nonspecific global translation
depression and apoptosis . However, this nonspecific
reaction can be circumvented by introduction of synthetic
siRNA [7,8], which can go into the RNAi pathway much like
the siRNAs produced from long dsRNA can. Alternatively,
RNAi may be triggered by a short hairpin RNA (shRNA)
synthesized from gene constructs with RNA polymerase III
(Pol III) or Pol II promoters . The Pol III–synthesized
shRNA is exported by exportin 5 from the nucleus to the
cytoplasm [10,11], where it is processed by Dicer to generate
siRNA, which goes into the RNAi pathway and mediates gene
silencing. The Pol II–expressed shRNA mimics micro RNAs
(miRNAs) and has one additional step of processing before
nuclear export: it is synthesized as a long Pol II transcript
[12,13] and is processed by the microprocessor complex that
contains Drosha and Pasha to produce the shRNA [14,15].
The simplicity and specificity of RNAi has made RNAi a
routine tool for investigation of gene functions in inverte-
brates and mammalian cells. Attempts have also been made to
develop RNAi as an in vivo reverse genetics tool in mice. An
early experiment directly injected long dsRNA into mouse
embryonic stem cells. Successful knockdown and the pheno-
type of the gene deletion were observed in embryogenesis
. This approach can produce only transient inhibition of
target gene expression. Another experiment used an oocyte-
specific promoter to express a long hairpin (;500 bp) against
c-Mos and demonstrated the expected phenotypes in oo-
genesis . Both approaches cannot be applied widely
because of the toxicity associated with long dsRNA in
differentiated somatic cells.
With the advent of Pol III promoter-directed synthesis of
shRNAs, several groups engineered transgenic mice using Pol
III–shRNA constructs [18–23]. In all cases, knockdown of the
marker transgene GFP in neonates was reported. In two cases,
phenotypes resembling genetic knockouts were observed in
developing mouse embryos  and hematopoietic stem cells
. Recently, Cre-lox–inducible Pol III promoters have been
demonstrated to knock down target genes and induce
phenotypes of gene deficiency in developing mouse embryos
[24–26]. However, the success rate of gene knockdown and
phenotypic expression is low using Pol III constructs [22,27];
despite the improvements using approaches employing
targeted insertion in embryonic stem cells and tetraploid
blastocyst complementation , widespread knockdown and
phenotypes of gene deletion that develop in postnatal
animals have not been demonstrated.
In addition to Pol III, Pol II can also direct shRNA synthesis
[29,30] and mediate efficient silencing in cultured cells .
Compared with Pol III promoters, Pol II promoter-directed
synthesis of shRNAs can be advantageous for transgenic
RNAi. Although limited choices of Pol III promoters have
been developed to express shRNA, a large repertoire of Pol II
promoters, including temporally and spatially specific and
inducible promoters, have been successfully used in trans-
genic mice. In addition, the current Pol III strategies rely on
the availability of various Cre transgenic mouse lines for
shRNA induction [24–26]. These Cre transgenic lines do not
exist in other mammalian species. Therefore, the application
of this Pol III strategy in other mammalian species is
currently impractical. Recent evidence indicates that miR-
NAs, the endogenous form of shRNAs, are downstream of Pol
II promoters  and are expressed by Pol II activity [32–34].
Therefore, strategies using Pol II–directed synthesis of shRNA
mimic the natural miRNA synthesis and could be an efficient
RNAi strategy in vivo.
To test this idea, we used a construct that is composed of a
human ubiquitin C promoter and an shRNA with the human
miRNA miR-30a structure  to generate transgenic mice.
We targeted the shRNA against the mouse Mn superoxide
dismutase (Sod2) gene, because Sod2-null mice generated by
standard knockout technology have striking postnatal phe-
notypes that have been well characterized [35–37]. This Pol II
strategy expressed siRNA widely in different tissues, knocked
down the Sod2 gene expression, and induced SOD2-hypo-
morphic phenotype. In addition, bigenic heterozygous mice
generated by crossing the two independent transgenic lines
showed phenotypes typical of the SOD2-null mouse, includ-
ing slow growth, fatty liver, dilated cardiomyopathy, and
premature death. Thus, this Pol II RNAi strategy can be an
alternative to gene knockout technology and opens the door
of RNAi to a wide array of Pol II transgenic strategies to
investigate gene functions in mice and other mammalian
The construct UbC-SOD2hp-EGFP (Figure 1; also see )
consists of the human ubiquitin C promoter, an shRNA-
coding hairpin placed in the first intron, followed by the
enhanced green fluorescent protein (EGFP) coding sequence
and a poly-adenylation signal. The shRNA-coding hairpin
mimics human microRNA miR-30a structure and target
mouse Sod2 mRNA. Using this construct, we generated
transgenic mice by pronuclear injection of fertilized eggs.
Screening of 62 founders yielded five positive lines (Figure
1A). Northern blots showed that two of the five lines (lines 8
and 26) expressed siRNA broadly (Figures 1B, 1C, and S1A)
and this expression was stable over multiple generations and
during aging (the same level of expression was observed in
animals from 40 to 200 d old; unpublished data). By real-time
PLoS Genetics | www.plosgenetics.org January 2006 | Volume 2 | Issue 1 | e100074
Transgenic RNAi by Pol II–Expressed shRNA
Reverse genetics studies gene functions by altering a gene and
observing the consequences. A powerful method of reverse
genetics in mammals is gene knockout by homologous recombina-
tion, which mutates a gene to prevent its functional expression.
Using this method, investigators have revealed the functions of
many genes. However, this method is relatively complex, time-
consuming, and costly. In addition, this method is limited to studies
in mice because it is not well established in other mammalian
species. The authors of this study tested an alternative method
using RNA interference (RNAi), which is a widely conserved
mechanism in eukaryotes and can mediate gene-specific silencing.
These investigators used RNA polymerase II (Pol II) to express a short
hairpin RNA (shRNA) that triggers destruction of the mRNA-
encoding Mn superoxide dismutase (SOD2) in transgenic mice.
These mice exhibit phenotypes that were typical in Sod2 knockout
mice, including elevated levels of oxidative stress in various tissues,
fat deposition in liver and muscles, dilated cardiomyopathy, and
premature death. These results open the door of RNAi to a wide
array of well-established Pol II transgenic strategies and offer a
technically simpler, cheaper, and quicker alternative to gene
knockout for reverse genetics in mice and other mammalian species.
PCR line 8 carried a single copy of the transgene while line 26
carried three copies (Figure S2). The transgene copy number
did not predict the level of siRNA levels since line 8 had a
higher level of siRNA than line 26 (see below).
The pattern of expression in transgenic mice differed from
the pattern in cultured cells in two regards: none of these
transgenic lines expressed detectable EGFP and shRNA (only
siRNA was detected). This contrasts with what we observed in
cultured cells, in which both were detectable . Never-
theless, the siRNA knocked down the abundance of SOD2 as
indicated by the decreased levels of the protein (Figures 1D
and S1B), mRNA (Figures 1E and S1C), and enzyme activity
(Figure 1F). The rapid processing of pre-miRNA in vivo
probably caused the lack of shRNA detection. This explan-
ation is consistent with the lack of pre-miRNA detection for
other endogenous miRNAs in mammalian cells . The lack
of EGFP was puzzling. One possibility was that the transgene
was altered. This was ruled out by sequencing the transgene
extracted by PCR, which revealed no alteration in the
transgene structure. Another possibility was that the pri-
miRNA processing in vivo was highly efficient, so that the
processing of pri-miRNA occurred before the splicing, and,
consequently, the mRNA could not be properly spliced and
exported to the cytoplasm for EGFP expression. The
processing of pri-miRNA involves Drosha . Therefore, if
our hypothesis was correct, inhibition of Drosha expression
should lead to EGFP expression. To test this, we transduced
fibroblasts isolated from the skeletal muscle of the transgenic
mice using a recombinant adenovirus that expresses an
shRNA against Drosha. By RT-PCR the Drosha mRNA levels
were substantially reduced (Figure S3A), indicating the
effectiveness of the shRNA. While the nontransduced cells
showed no detectable EGFP fluorescence (Figure S3B and
S3C), the transduced cells expressed EGFP (Figure S3D and
S3E). This result supports our hypothesis.
To confirm the knockdown of Sod2 gene expression, we
examined the in vivo consequence of SOD2 deficiency. First,
we compared the activity of the mitochondrial enzyme
succinate dehydrogenase (SDH) between the transgenic and
wild-type animals, because SOD2 knockout is known to cause
a decrease in SDH activity [37,38]. We stained tissue sections
from the heart using histochemical staining, and observed a
decrease in the staining intensity in line 8 (Figure 2A and 2B).
Second, we isolated fibroblasts from skeletal muscle and
measured the levels of superoxide in these cells. We observed
that the superoxide levels were increased in the two trans-
genic lines (Figure 2C). This increase was higher in line 8 than
in line 26, and, therefore, was correlated with the degree of
SOD2 knockdown (see below). Third, we tested the sensitivity
of the fibroblasts to oxidative stress induced by t-butylhy-
droperoxide (t-BuOOH) treatment. Fibroblasts from trans-
genic line 8 showed a higher sensitivity than the wild-type
Figure 1. shRNA Expression and Knockdown of Sod2 Gene Expression
(A) Schematic illustration of the transgene construct. The shRNA was
designed to mimic human miR-30a structure (for details, see ).
(B) PCR analysis of tail DNA identified transgenic founders. Cþ indicates
positive control; C?, negative control. Numbers indicate examples of
various transgenic lines.
(C) Northern blots detected shRNA expression in transgene-positive line-
8 mice, but not in line-60 mice. Total RNA (30 lg) was loaded in each
lane. The tissues are lung (1), heart (2), skeletal muscle (3), kidney (4), liver
(5), brain (6), stomach (7), and spleen (8). Cþis the siRNA-positive control.
(D) Western blots of SOD2 protein compare the SOD2 levels in the above
tissues between line-8 mice and wild-type mice. Due to different levels of
SOD2 in different tissues, different amounts of total protein from
different tissues had to be loaded in order to maintain the assay in linear
range. The amounts of proteins are the following in micrograms: (1) 30;
(2–4) 10; (5–6) 15; (7) 20; and (8) 40. þ indicates transgene positive; ?,
(E) SOD2 mRNA levels in the above tissues from transgenic line-8 mice
measured by real-time PCR (n¼4; all ‘‘n’’ indicates mouse numbers). The
levels were normalized to the level of SOD2 mRNA in tissues from the
wild-type littermates, which were set as 100% (column C).
(F) Levels of SOD2 activity in tissue lysates of transgenic line-8 mice
compared with the wild-type littermates (n ¼ 4).
PLoS Genetics | www.plosgenetics.org January 2006 | Volume 2 | Issue 1 | e100075
Transgenic RNAi by Pol II–Expressed shRNA
cells (Figure 2D), consistent with the lowered levels of SOD2
protein in these cells (Figure 2E). To ensure that the
enhanced sensitivity was caused by the knockdown of SOD2
expression, we transduced the fibroblasts with a recombinant
adenovirus that expressed a Sod2 gene whose mRNA was
resistant to the siRNA expressed by the transgene. Expression
of this SOD2 molecule (Figure 2E) rescued the cells from their
hypersensitivity to oxidative stress (Figure 2D).
Based on these data, we conclude that ubiquitin C
promoter–directed shRNA synthesis effectively silenced the
target molecule in transgenic mice. Despite this significant
knockdown (by 60%–90%) in all the tissues examined and the
evidence of functional SOD2 deficiency, the two transgenic
lines were viable to 400 d (observed to date). SOD2-null
phenotypes, including small body size, dilated cardiomyop-
athy, lipid deposition in liver and heart, and premature
death, were not observed. To determine whether knockout
phenotypes could be generated, we crossed the two lines that
expressed the siRNA to generate bigenic heterozygous trans-
genic mice. We took this approach because it was advanta-
geous compared with generating homozygous animals of each
lines, the phenotype of which could be complicated by the
potential gene disruption at the transgene insertion site.
The line 8/26 bigenic mice expressed a higher level of
siRNA than either of the singly transgenic lines (Figure 3A),
leading to knockdown of SOD2 protein and mRNA to nearly
undetectable levels (Figure 3B and 3C). These mice exhibited
phenotypes similar to the previously reported SOD2 knock-
out mice , including smaller body size than the wild-type
littermates (Figure 4A) and death within 20 d after birth (in
34 bigenic animals that we obtained so far). In addition, they
developed dilated cardiomyopathy (Figure 4B) and had
increased lipid deposition in the heart (Figure 4C and 4D)
and the liver (Figure 4E and 4F).
Several studies have shown that some shRNA or siRNA
could trigger interferon response [39–41]. This raises the
possibility that interferon response might be responsible for
the phenotype observed in these mice. To test this, we
examined the levels of two genes known to be dramatically
induced by the dsRNA-triggered interferon response, 29,59-
oligoadenylate synthetase 1 (OSA1), and signal transducer and
activator of transcription 1 (STAT1) [40,41]. Real-time PCR
analysis of the mRNA of OSA1 and STAT1 failed to detect
changes in the expression levels of these two genes in the
heart, spleen, and liver in the transgenic line 8 (Figure 5),
indicating no interferon response in these transgenic mice.
Taken together, the phenotypes are likely caused by the
specific effect of SOD2 knockdown because (1) the siRNA was
expressed widely; (2) the consequences of SOD2 deficiency
were observed in the transgenic mice; (3) in cells isolated
from these mice the hypersensitivity to oxidative stress was
corrected by the Sod2 gene that was resistant to the siRNA; (4)
Figure 2. Consequence of the SOD2 Knockdown
(A) Histochemical staining reveals that SDH activity in the heart of transgenic line-8 mice was reduced compared with the wild-type littermates (B).
(C) ROS levels are increased in fibroblasts from the skeletal muscles of transgenic line-26 (top panel) and line-8 (bottom panel) mice, compared with
those from the wild-type mice (middle panel). AU, arbitrary units.
(D) Fibroblasts from the transgenic line-8 mice have elevated sensitivity to oxidative stress compared with those from the wild-type mice, and this
sensitivity can be corrected by expressing an RNAi-resistant SOD2 (line 8 þSOD2r). The data are means observed in cells isolated from four individual
mice. Error bars are SEM. The asterisks indicate significant difference as compared to either WT or rescued cells (p , 0.05).
(E) Western blot detects SOD2 protein levels in fibroblasts isolated from the skeletal muscle of the wild-type and transgenic line-8 mice. The third lane is
from the line-8 cells transduced with RAd expressing the siRNA-resistant SOD2.
PLoS Genetics | www.plosgenetics.orgJanuary 2006 | Volume 2 | Issue 1 | e10 0076
Transgenic RNAi by Pol II–Expressed shRNA
the phenotype typical of the SOD2-null mouse was observed
when siRNA levels were increased in the bigenic 8/26 mice;
and (5) levels of OSA1 and STAT1, two molecules involved in
the dsRNA-induced interferon response, were unchanged.
These results demonstrate that Pol II–mediated expression of
shRNA in transgenic mice can be used to investigate gene
functions in mammals.
Thus, the Pol II–directed synthesis of shRNA can be an
alternative to gene knockout technology for reverse genetics
in mammals. Although gene knockout remains a useful
approach for the complete gene deletion or gene modifica-
tion, our RNAi approach can achieve near knockout
conditions and is economical in cost and time. In addition,
the construct design is simple and in principle not different
from the standard transgene design for gene overexpression.
The placement of the shRNA-encoding hairpin is flexible: it
can be placed in introns or in the 39 untranslated regions
[13,42], and the shRNA-encoding transcript is not required to
encode a protein [13,29,30]. Therefore, a wide variety of
transgene promoters that have already been successfully used
to overexpress genes in transgenic mice can be readily
adapted for suppressing the genes. When spatially and
temporally specific promoters are used, controlled suppres-
sion of the target gene can be achieved. Furthermore, this
approach provides a more flexible system to model hypo-
morphic allele function. While the standard knockout
technology reduces the target gene expression to 50% or
0% in most instances, transgenic RNAi can reduce the target
gene to more variable degrees in different lines of transgenic
mice or by induction using small molecules in transgenic
mice made with inducible Pol II promoters [43,44]. The
simplicity of this approach can accelerate the generation of
models for diseases that are caused by various degrees of
genetic hypomorphism, such as autosomal dominant poly-
cystic kidney disease , various cancers [46,47], and other
diseases. Finally, gene knockout by homologous recombina-
tion has not been established in other mammalian species.
Current Pol III strategies rely on the crosses with numerous
Cre-expression transgenic mouse lines [24–26], which are not
available in other mammalian species. Our knockdown
strategy can overcome these limitations and be used to carry
out reverse genetics and to generate disease models in other
Materials and Methods
Generation of SOD2 knockdown transgenic mice. The transgene
construct that contains the hairpin targeting the Sod2 gene under the
control of human ubiquitin C promoter (UbC-SOD2hp-EGFP) has
been described previously . The transgenic mice were made by
pronuclear injection of the linearized construct into the fertilized
eggs, which were generated from crossings of C57BL/6 and SJL.
Positive founders and offspring carrying the transgene were
identified by PCR of the tail DNA using the primers 59-
CGGCGCGGGTCTTGTAGTTGC-39 (reverse). The transgenic lines
were maintained by crossing founders to C57BL/6. To generate
bigenic heterozygous transgenic mice, two shRNA-expressing lines
were crossed and the doubly transgenic mice were identified by
quantitative real-time PCR of tail DNA.
Northern blot. Mice were decapitated under anesthesia, and
various tissues were quickly dissected, snap-frozen in liquid nitrogen,
and stored at?80 8C. The total RNA was extracted from frozen mouse
tissues using Trizol (Sigma, St Louis, Missouri, United States). Thirty
micrograms of total RNA was fractionated on 15% polyacrylamide
gels and transferred onto Hybond TM-Nþ membrane (Amersham
the membrane was probed with32P-labeled synthetic RNA oligonu-
cleotide complementary to the antisense strand of the mouse Sod2
shRNA as described previously . For some blots, the membranes
were reprobed with32P-labeled synthetic DNA oligonucleotide (59-
ACGAATTTGCGTGTCATCCTTGCG-39) complementary to mouse
Western blot. The frozen mouse tissues were homogenized in ice-
cold lysis buffer containing 0.4% NP-40, 0.2 mM Na3VO4, 20 mM
HEPES (pH 7.9), and a cocktail of protease inhibitors (Complete-
Mini; Sigma). The protein content in the cleared lysate was
determined using the BCA assay. Equal amount of total proteins
from transgenic and wild-type control animals was resolved by 15%
SDS-PAGE and blotted onto GeneScreen Plus membrane (Perki-
nElmer, Wellesley, Massachusetts, United States). Proteins were
detected using specific primary antibodies and the SuperSignal kit
(Pierce Biotechnology, Rockford, Illinois, United States) and photo-
graphed using the Kodak Digital Image Station 440CF. The primary
antibodies were: rabbit anti-Mn superoxide dismutase (SOD2; 1:1,000,
Stressgen Biotechnologies, San Diego, California, United States) and
mouse anti–glyceraldehye-3-phosphate dehydrogenase (GAPDH;
1:10,000; Abcam, Cambridge, United Kingdom). After detection of
SOD2, the membrane was stripped for 30 min at 55 8C in a buffer
containing 100 mM b-mercaptoethanol, 2% SDS (w/v), and 62.5 mM
Figure 3. The Level of SOD2 Expression Was Knocked Down Further in
Bigenic Transgenic Mice Generated by Crossing the Two Lines (Lines 8
(A) Northern blots indicate that siRNA levels are further increased in the
8/26 bigenic mice.
(B) Western blots demonstrate that SOD2 protein levels are further
knocked down in the 8/26 bigenic mice.
(C) Real-time PCR shows that SOD2 mRNA levels are further lowered in
the 8/26 bigenic mice.
PLoS Genetics | www.plosgenetics.org January 2006 | Volume 2 | Issue 1 | e100077
Transgenic RNAi by Pol II–Expressed shRNA
Tris-HCl (pH 6.7), and used again for detection of GAPDH
immunoreactivity, which served as loading control.
Real-time quantitative RT-PCR. Total RNA isolated using the TRI
reagent was further purified with the RNeasy kit (Qiagen, Valencia,
California, United States) and subjected to digestion on column with
RNase-Free DNase (Qiagen). One microgram of purified total RNA
from each sample was reversely transcribed to cDNA with oligo-dT
primer using the RT kit (Invitrogen, Carlsbad, California, United
States). The cDNA was used for quantitative PCR with SYBR green kit
(Qiagen) according to manufacturer’s instruction. The primer
concentration was 500 nM. Cycling conditions were 15 min at 95 8C
(to activate the hot-start Taq polymerase supplied with the SYBR
Green detection kit), followed by 40 cycles of 15 s at 94 8C, 30 s at 60
8C, and 20 s at 72 8C. During amplification the fluorescence signal,
which is proportional to the amount of dsDNA produced, was
monitored. A complete amplification profile for each of the 96 wells
of a PCR plate was obtained, which was used for the analysis. At the
end of the PCR run, melting curves of the amplified products were
obtained, which were used to determine the specificity of the
amplification reaction. In pilot experiments, aliquots of the amplified
products were separated on 3% agarose gels to ensure amplification
of specific products of the predicted length. The amplification curves
were used to calculate the threshold cycle number at which the
amplification curve reaches the beginning of the linear phase of
amplification. The threshold cycle number for Sod2 gene was
normalized to those of the housekeeping genes GAPDH and
ribosomal RNA L17. The knockdown of the Sod2 expression was
determined by calculating the fold change of Sod2 in transgenic
tissues relative to the wild-type tissues.
SOD2 activity. SOD2 activity was determined by inhibition of
xanthine/xanthine oxidase–induced cytochrome C reduction .
Frozen tissues were homogenized in 10 volumes of ice-cold buffer (10
mM KH2PO4[pH 7.4], 20 mM EDTA, 30 mM KCl). The supernatant
was collected after centrifugation of homogenates in a desktop
microcentrifuge for 15 min at 4 8C and measured for protein
concentration using BAC assay (Pierce). The activity of SOD1 was
inactivated by treating the lysates with 5 mM KCN before SOD2
activity assay. To a cuvette, 967 ll of solution A (50 lM xanthine, 20
lM acetylated cytochrome C, 25 lM KH2PO4, 25 lM Na2HPO4, and
0.1 mM EDTA) was added, followed by 16.7 ll of the treated lysates
and 16.7 ll of solution B (0.2 U/ml xanthine oxidase in 0.1 mM EDTA).
The absorbance at 550 nm was read at 1-min intervals for 10 min.
SOD2 activity in knockdown lysates was determined by comparing
with an SOD2 standard curve (SOD2 enzyme; Sigma) and expressed
relative to that in the wild-type lysates.
SDH histochemistry. Immediately following death, tissues were
harvested and frozen on dry ice and sectioned at 20 lm thickness at
?25 8C. SDH staining was performed and evaluated as previously
described on frozen sections .
Evaluation of ROS levels and tolerance to oxidative stress in
fibroblasts. Fibroblasts were isolated from skeletal muscle in 15-d-old
mice using the method modified from Crisona et al. . Skeletal
Figure 4. The 8/26 Bigenic Mice Display the Phenotype of a SOD2-Null Animal
(A) Retarded growth (7-d-old animals).
(B) Dilated cardiomyopathy (H&E-stained coronal sections of heart). LV, left ventricle; RV, right ventricle.
(C–F) Lipid deposition in heart (C and D) and liver (E and F) stained with Oil Red O.
PLoS Genetics | www.plosgenetics.orgJanuary 2006 | Volume 2 | Issue 1 | e10 0078
Transgenic RNAi by Pol II–Expressed shRNA
muscles were removed from hindlimbs and minced in Hank’s
balanced salt solution. The muscle slurry was digested with
collagenase type I (0.5% in Hank’s balanced salt solution; Invitrogen)
at 37 8C for 1 h. The cells were pelleted by centrifugation and
subjected to further digestion with trypsin at 37 8C for 1 h. After
another round of centrifugation and washing, the cells were plated to
75-cm2flasks and grown in DMEM supplemented with 10% FBS, 100
mg/ml streptomycin, and 100 U/ml penicillin. The cultured cells were
split every 3 d and used before four passages. For determination of
ROS (superoxide), 500,000 primary (passage 2) wild-type and Sod2
shRNA transgenic skeletal muscle fibroblasts were incubated in
DMEM media containing 10% fetal calf serum, antibiotics, and 2 lM
dihydroethidine (Molecular Probes, Eugene, Oregon, United States)
for 30 min at 37 8C in the dark. After incubation, the cells were
centrifuged, resuspended in freshly prepared cold FACS staining
buffer (PBS, 1% BSA), and the FL-2 fluorescence of 20,000 cells per
sample was analyzed using a Becton Dickinson LSR cytometer (Palo
Alto, California, United States). In the same experiments, passage-1
wild-type and Sod2?/?mouse embryonic fibroblasts were included as
controls. To prevent bias, the ROS determination was performed in a
blind fashion. For viability assay, the cells were plated to 96-well
plates at 1 3 104cells per well and allowed to attach overnight at 37
8C. On the following day, the cells were treated with t-BuOOH at
different concentrations for 2 h. After three washes with PBS, the
cells were grown in DMEM containing 10% FBS and MTS reagent
(Promega, Madison, Wisconsin, United States) for 2 h. To measure the
viability, the absorbance of cultures was read at 495 nm with a plate
reader and normalized to the average absorbance of the cells
untreated with tBuOOH. For the rescue experiments, the transgenic
cells (two passages) were infected in 25-cm2flasks with adenoviral
vectors expressing RNAi-resistant Sod2 gene (see below). After 3 d, the
transduced cells were split into 96-well plates, grown for 24 h, and
treated with t-BuOOH; the viability was measured as described above.
Statistical analysis was performed using ANOVA followed by Tukey
post-hoc test to compare group means.
Adenoviral vector expressing RNAi-resistant Sod2. ViraPower
adenoviral kit (Invitrogen) was used. An RNAi-resistant Sod2 gene
was generated by introducing four silent mutations within shRNA
target region. The target sequence of SOD2 mRNA was 59-
AAGGGAGATGTTACAACT-39, and the target sequence of RNAi-
resistant SOD2 mRNA was 59-AAG GGT GAC GTA ACT ACT-39
(bold letters indicate the silent mutations). The CMV-SOD2 expres-
sion cassette was introduced into ViraPower adenoviral vector (RAd;
Invitrogen) using the Gateway method. The RAd-SOD2 vector was
transiently transfected into a RAd producer cell line 293A that stably
expresses E1 proteins required for production of adenovirus. The
RAd-SOD2 was harvested from transiently transfected 293A cells and
amplified by infecting fresh 293A cells. The virus was purified by CsCl
gradient centrifugation, dialyzed against a buffer containing 10 mM
Tris (pH 7.5), 1 mM MgCl2, and 10% glycerol, and stored in aliquots
at?80 8C. The viral titer (plaque-forming unit) was determined using
293A cells according to the manufacturer’s instructions. To transduce
fibroblast cells, the virus was used at a multiplicity of infection (MOI;
50) and incubated with the culture for 3 h before the media were
changed. The transduced cells were used for viability assay at 48 h
Oil Red O staining. Oil Red O staining was performed on frozen
sections. The fresh tissues collected as described above were frozen in
powdered dry ice. The frozen sections (12 lm) were cut using a
Cryostat and stained with Oil Red O solution on slides as follows: the
sections were fixed in 10% formalin for 5 min, washed with several
changes of phosphate-buffered saline, stained in prewarmed Oil Red
O solution (60 8C; Sigma) for 8 min, washed again several times with
distilled water, stained in Gill’s hematoxylin solution for 30 s, and
washed several times in distilled water. After mounted with cover-
slips, the stained sections were observed under microscope and
Figure S1. shRNA Expression and Knockdown of Sod2 Gene
Expression in Line 26
(A) Northern blots detect shRNA expression in transgene-positive
line 26. Total RNA (30 lg) was loaded in each lane. The tissues are
lung (1), heart (2), skeletal muscle (3), kidney (4), liver (5), brain (6),
stomach (7), and spleen (8).
(B) Western blots compare the SOD2 protein levels in the above
tissues between line-26 mice and wild-type mice. þ indicates trans-
gene positive, and ? indicates transgene negative. The amounts of
proteins are loaded in the same order as described in Figure 2.
(C) SOD2 mRNA levels in the above tissues from transgenic line 26
measured by real-time PCR (n¼4). The levels were normalized to the
level of SOD2 mRNA in tissues from the wild-type littermates, which
were set as 100% (column C).
Found at DOI: 10.1371/journal.pgen.0020010.sg001 (46 KB PDF).
Figure S2. Determining the Copy Number of the Transgene by Real-
(A) Testing the specificity of the primers used for real-time PCR.
HEK293 cells were used as reference for gene copy number of
ubiquitin C. NT, nontransgenic.
(B) Estimation of UbC-SOD2hp-EGFP gene copy numbers. A 118-bp
segment in human ubiquitin C promoter was amplified using a pair of
specific primers. Also amplified was human and mouse SOD1 gene
using a pair of primers that are complementary to both genes. Both
Ubiquitin C and SOD1 are single-copy genes. The threshold cycle
number value of Ubiquitin C was normalized against the SOD1
detected in the same sample. The normalized value from HEK293
genomic DNA represents two copies of the ubiquitin C gene. By
normalizing this value from mouse genomic DNA samples against the
value from HEK293 cells, the estimates of copy numbers of the
ubiquitin C transgene were obtained. The PCR primers used for
human and mouse Sod1 gene were 59-GACCTGGGCAATGT
GACTGCTG-39 (forward) and 59-CACCAGTGTACGGCCAAT
GATG-39 (reverse); and for ubiquitin C promoter were 59-
TTTCCTCGCCTGTTCCGCTC-39 (reverse). Bars are averages from
four to six animals.
Found at DOI: 10.1371/journal.pgen.0020010.sg002 (33 KB PDF).
Figure S3. Silencing Drosha Unblocked the Expression of EGFP
Transgene in Muscle Fibroblasts
(A) Drosha was knocked down by RNAi in the fibroblasts from wild-
type and line-8 transgenic mice. The cells were transduced with
adenoviral vectors expressing an shRNA against mouse Drosha
(shRNA stem sequence: 59-GGATGAAGATTTAGAGAGTTC-39). Four
days after transduction, the total RNA was extracted from the cells
and used for RT-PCR to detect Drosha. The ribosomal RNA L17 was
magnified in parallel as input control. The PCR was run for 28 cycles
using the following primers for Drosha: 59-GAGCCTAGAGGAAGC
CAAACA-39 (forward) and 59-GCCGGACGTGAGTGAAGAT-39 (re-
verse); for L17: 59-CGGTATAATGGTGGAGTTG-39 (forward) and 59-
(B) No EGFP fluorescence could be detected in fibroblasts isolated
from line-8 transgenic mice.
(C) The same field as in (B) was stained with DAPI.
(D) Three days after the fibroblasts were transduced with an
adenoviral vector that expressed an shRNA against mouse Drosha,
EGFP fluorescence was detected.
(E) The same field as in (D) was stained with DAPI.
Found at DOI: 10.1371/journal.pgen.0020010.sg003 (48 KB PDF).
Figure 5. In Vivo Expression of SOD2 shRNA with Authentic miR30a
Structure Did Not Up-Regulate the Expression of OSA1 and STAT1
Levels of the mRNAs were determined by real-time PCR. The levels of
OAS1 and STAT1 mRNAs in the shRNA transgenic tissues from line-8 mice
were normalized to GAPDH mRNA and expressed relative to that of wild-
type littermate tissues. Data represent means from four mice þ SEM.
PLoS Genetics | www.plosgenetics.orgJanuary 2006 | Volume 2 | Issue 1 | e100079
Transgenic RNAi by Pol II–Expressed shRNA
Acknowledgments Download full-text
We thank members of the Xu lab and Dr. Alonzo Ross for advice and
support, Phillip Zamore for discussion and commenting on the
manuscript, and Steve Jones and the University of Massachusetts
Medical School transgenic core for pronuclear injection. This work
was supported by grants from the Amyotrophic Lateral Sclerosis
(ALS) Association, National Institutes of Health (NIH)/National
Institute of Neurological Disorders and Stroke (NINDS)
(R01NS048145), NIH/National Institute on Aging (NIA)
(R21AG023808), and The Robert Pachard Center for ALS Research
at Johns Hopkins to ZX, and NIH/NIA (RO1AG18679) to SM. The
contents of this report are solely the responsibility of the authors and
do not necessarily represent the official views of the NIH.
Author contributions. XGX, HZ, and ZX conceived and designed
the experiments. XGX, HZ, ES, and SM performed the experiments.
XGX, HZ, ES, SM, and ZX analyzed the data. XGX, HZ, ES, and SM
contributed reagents/materials/analysis tools. ZX wrote the paper.
Competing interests. XGX, HZ, and ZX are authors of a pending
patent on the Pol II–shRNA construct used in this study.
1. Austin CP, Battey JF, Bradley A, Bucan M, Capecchi M, et al. (2004) The
knockout mouse project. Nat Genet 36: 921–924.
2. Hannon GJ, Rossi JJ (2004) Unlocking the potential of the human genome
with RNA interference. Nature 431: 371–378.
3.Mello CC, Conte D (2004) Revealing the world of RNA interference. Nature
4.Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, et al. (1998) Potent
and specific genetic interference by double-stranded RNA in Caenorhabditis
elegans. Nature 391: 806–811.
5. Tomari Y, Zamore PD (2005) Perspective: Machines for RNAi. Genes Dev
6.Gil J, Esteban M (2000) Induction of apoptosis by the dsRNA-dependent
protein kinase (PKR): Mechanism of action. Apoptosis 5: 107–114.
7. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, et al. (2001)
Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured
mammalian cells. Nature 411: 494–498.
8.Caplen NJ, Parrish S, Imani F, Fire A, Morgan RA (2001) Specific inhibition
of gene expression by small double-stranded RNAs in invertebrate and
vertebrate systems. Proc Natl Acad Sci U S A 98: 9742–9747.
9. Shi Y (2003) Mammalian RNAi for the masses. Trends Genet 19: 9–12.
10. Lund E, Guttinger S, Calado A, Dahlberg JE, Kutay U (2004) Nuclear export
of microRNA precursors. Science 303: 95–98.
11. Yi R, Qin Y, Macara IG, Cullen BR (2003) Exportin-5 mediates the nuclear
export of pre-microRNAs and short hairpin RNAs. Genes Dev 17: 3011–
12. Zeng Y, Wagner EJ, Cullen BR (2002) Both natural and designed micro
RNAs can inhibit the expression of cognate mRNAs when expressed in
human cells. Mol Cell 9: 1327–1333.
13. Zhou H, Xia XG, Xu Z (2005) An RNA polymerase II construct synthesizes
short-hairpin RNA with a quantitative indicator and mediates highly
efficient RNAi. Nucl Acids Res 33: e62.
14. Lee Y, Ahn C, Han J, Choi H, Kim J, et al. (2003) The nuclear RNase III
Drosha initiates microRNA processing. Nature 425: 415–419.
15. Denli AM, Tops BBJ, Plasterk RHA, Ketting RF, Hannon GJ (2004)
Processing of primary microRNAs by the Microprocessor complex. 432:
16. Wianny F, Zernicka-Goetz M (2000) Specific interference with gene
function by double-stranded RNA in early mouse development. Nat Cell
Biol 2: 70–75.
17. Stein P, Svoboda P, Schultz RM (2003) Transgenic RNAi in mouse oocytes:
A simple and fast approach to study gene function. Dev Biol 256: 187–193.
18. Kunath T, Gish G, Lickert H, Jones N, Pawson T, et al. (2003) Transgenic
RNA interference in ES cell-derived embryos recapitulates a genetic null
phenotype. Nat Biotechnol 21: 559–561.
19. Hemann MT, Fridman JS, Zilfou JT, Hernando E, Paddison PJ, et al. (2003)
An epi-allelic series of p53 hypomorphs created by stable RNAi produces
distinct tumor phenotypes in vivo. Nat Genet 33: 396–400.
20. Hasuwa H, Kaseda K, Einarsdottir T, Okabe M (2002) Small interfering
RNA and gene silencing in transgenic mice and rats. FEBS Lett 532: 227–
21. Rubinson DA, Dillon CP, Kwiatkowski AV, Sievers C, Yang L, et al. (2003) A
lentivirus-based system to functionally silence genes in primary mamma-
lian cells, stem cells and transgenic mice by RNA interference. Nat Genet
22. Carmell MA, Zhang L, Conklin DS, Hannon GJ, Rosenquist TA (2003)
Germline transmission of RNAi in mice. Nat Struct Biol 10: 91–92.
23. Tiscornia G, Singer O, Ikawa M, Verma IM (2003) A general method for
gene knockdown in mice by using lentiviral vectors expressing small
interfering RNA. Proc Natl Acad Sci U S A 100: 1844–1848.
24. Ventura A, Meissner A, Dillon CP, McManus M, Sharp PA, et al. (2004) Cre-
lox-regulated conditional RNA interference from transgenes. Proc Natl
Acad Sci U S A 101: 10380–10385.
25. Chang HS, Lin CH, Chen YC, Yu WCY (2004) Using siRNA technique to
generate transgenic animals with spatiotemporal and conditional gene
knockdown. Am J Pathol 165: 1535–1541.
26. Coumoul X, Shukla V, Li C, Wang RH, Deng CX (2005) Conditional
knockdown of Fgfr2 in mice using Cre-LoxP induced RNA interference.
Nucl Acids Res 33: e102.
27. Xia XG, Zhou H, Xu ZS (2005) Promises and challenges in developing RNAi
as a research tool and therapy for neurodegenerative diseases. Neuro-
degenerative Dis 2: In press.
28. Seibler J, Kuter-Luks B, Kern H, Streu S, Plum L, et al. (2005) Single copy
shRNA configuration for ubiquitous gene knockdown in mice. Nucl Acids
Res 33: e67.
29. Zeng Y, Cullen BR (2003) Sequence requirements for micro RNA
processing and function in human cells. RNA 9: 112–123.
30. Xia H, Mao Q, Paulson HL, Davidson BL (2002) siRNA-mediated gene
silencing in vitro and in vivo. Nat Biotechnol 20: 1006–1010.
31. Lagos-Quintana M, Rauhut R, Meyer J, Borkhardt A, Tuschl T (2003) New
microRNAs from mouse and human. RNA 9: 175–179.
32. Cai X, Hagedorn C, Cullen BR (2004) Human microRNAs are processed
from capped, polyadenylated transcripts that can also function as mRNAs.
RNA 10: 1957–1966.
33. Bracht J, Hunter S, Eachus R, Weeks P, Pasquinelli AE (2004) Trans-splicing
and polyadenylation of let-7 microRNA primary transcripts. RNA 10: 1586–
34. Lee Y, Kim M, Han J, Yeom KH, Lee S, et al. (2004) MicroRNA genes are
transcribed by RNA polymerase II. EMBO J 23: 4051–4060.
35. Li Y, Huang TT, Carlson EJ, Melov S, Ursell PC, et al. (1995) Dilated
cardiomyopathy and neonatal lethality in mutant mice lacking manganese
superoxide dismutase. Nat Genet 11: 376–381.
36. Huang TT, Carlson EJ, Raineri I, Gillespie AM, Kozy H, et al. (1999) The use
of transgenic and mutant mice to study oxygen free radical metabolism.
Ann NY Acad Sci 893: 95–112.
37. Melov S, Coskun P, Patel M, Tuinstra R, Cottrell B, et al. (1999)
Mitochondrial disease in superoxide dismutase 2 mutant mice. Proc Natl
Acad Sci U S A 96: 846–851.
38. Hinerfeld D, Traini MD, Weinberger RP, Cochran B, Doctrow SR, et al.
(2004) Endogenous mitochondrial oxidative stress: Neurodegeneration,
proteomic analysis, specific respiratory chain defects, and efficacious
antioxidant therapy in superoxide dismutase 2 null mice. J Neurochem 88:
39. Hornung V, Guenthner-Biller M, Bourquin C, Ablasser A, Schlee M, et al.
(2005) Sequence-specific potent induction of IFN-alpha by short interfer-
ing RNA in plasmacytoid dendritic cells through TLR7. Nat Med 11: 263–
40. Bridge AJ, Pebernard S, Ducraux A, Nicoulaz AL, Iggo R (2003) Induction
of an interferon response by RNAi vectors in mammalian cells. Nat Genet
41. Pebernard S, Iggo RD (2004) Determinants of interferon-stimulated gene
induction by RNAi vectors. Differentiation 72: 103–111.
42. Ling X, Li F (2004) Silencing of antiapoptotic survivin gene by multiple
approaches of RNA interference technology. Biotechniques 36: 450–454,
43. Ryding AD, Sharp MG, Mullins JJ (2001) Conditional transgenic technol-
ogies. J Endocrinol 171: 1–14.
44. Bockamp E, Maringer M, Spangenberg C, Fees S, Fraser S, et al. (2002) Of
mice and models: Improved animal models for biomedical research. Physiol
Genomics 11: 115–132.
45. Lantinga-van Leeuwen IS, Dauwerse JG, Baelde HJ, Leonhard WN, van de
Wal A, et al. (2004) Lowering of Pkd1 expression is sufficient to cause
polycystic kidney disease. Hum Mol Genet 13: 3069–3077.
46. Gaspar C, Fodde R (2004) APC dosage effects in tumorigenesis and stem
cell differentiation. Int J Dev Biol 48: 377–386.
47. Moynahan ME (2002) The cancer connection: BRCA1 and BRCA2 tumor
suppression in mice and humans. Oncogene 21: 8994–9007.
48. Asimakis GK, Lick S, Patterson C (2002) Postischemic recovery of
contractile function is impaired in SOD2þ/? but not SOD1þ/? mouse
hearts. Circulation 105: 981–986.
49. Crisona NJ, Allen KD, Strohman RC (1997) Muscle satellite cells from
dystrophic (mdx) mice have elevated levels of heparan sulphate proteogly-
can receptors for fibroblast growth factor. J Muscle Res Cell Motil 19: 43–
PLoS Genetics | www.plosgenetics.org January 2006 | Volume 2 | Issue 1 | e100080
Transgenic RNAi by Pol II–Expressed shRNA